Do they really think we’re just going to forget?

Detailed Analysis

A Reddit post titled "Do they really think we're just going to forget?" captures a growing undercurrent of customer frustration directed at Anthropic's communication practices around product commitments and pricing changes. The post centers on two specific grievances: first, an apparent unfulfilled promise regarding "Fable 5" being included in subscription plans, with the poster demanding Anthropic either confirm or deny whether this remains the plan rather than letting the issue quietly fade from public discussion; and second, a pricing maneuver described as a "50% increase, then 50% decrease" on weekly usage limits, which the author characterizes as manipulative rather than transparent. The tone of the post—frustrated, accusatory, and skeptical of corporate silence—reflects a broader pattern of user sentiment that has emerged as Anthropic has scaled Claude's commercial offerings, particularly around Claude Code and subscription tiers.

This complaint matters because it touches on a core tension facing AI companies as they mature from research labs into full-fledged commercial platforms: the gap between how they communicate with technical/research audiences versus how they manage expectations with paying retail customers. Anthropic, like OpenAI, has increasingly rolled out usage limits, tiered pricing, and subscription restructuring as compute costs and demand fluctuate. When such changes are announced without full transparency—or when previously stated roadmap items go unmentioned rather than formally updated—users interpret this as evasiveness rather than the practical realities of iterating pricing models. The specific complaint about limits being raised and then lowered suggests a pattern where companies may test aggressive changes, gauge backlash, and partially walk them back without directly acknowledging the original overreach, a tactic that erodes trust even when the net effect might seem neutral or even user-favorable.

The stakes are elevated by the sentence at the end of the post: the author explicitly ties customer retention to competitive pressure from OpenAI, suggesting that Anthropic's goodwill with developers and power users is not unconditional loyalty but rather contingent on Claude Code remaining functionally superior or uniquely positioned. This reflects the broader competitive landscape in 2025–2026, where the "moat" for AI coding assistants and agentic tools is increasingly thin — technical capabilities converge quickly across providers, and switching costs for developers are relatively low compared to consumer-facing products. In this environment, trust and communication style become differentiators nearly as important as raw model performance, since frustrated power users have credible alternatives and are quick to signal potential defection publicly.

More broadly, this incident is emblematic of a recurring theme in the AI industry: companies built on research-lab cultures (opaque, iterative, hypothesis-driven) often struggle when those same instincts are applied to product and pricing decisions, where customers expect the transparency and reliability associated with mature SaaS businesses. Anthropic's positioning as a safety-focused, mission-driven company has cultivated a user base that holds it to a higher standard of honesty and consistency than it might apply to a more purely profit-driven competitor. Community backlash of this kind—amplified through platforms like Reddit—serves as an informal but increasingly important feedback loop that pressures AI companies to formalize communication about roadmap changes, pricing adjustments, and walked-back promises, or risk cultivating exactly the kind of cynical, retention-eroding sentiment expressed in this post.